Memory

In translation and localization, memory typically refers to translation memory (TM), a database that stores previously translated text segments and their corresponding translations. These stored translations can be reused when the same or similar content appears again.

Memory helps translators avoid repeating work while maintaining consistent language across projects.

How Translation Memory Works

Translation memory systems store and retrieve translated segments during the translation process.

Segment Storage Each source segment and its translated equivalent are saved as a pair in the translation memory database.

Match Detection When similar content appears in new projects, the system identifies matches in the translation memory.

Translation Suggestions Previously translated segments are suggested to translators for reuse or modification.

Continuous Growth The translation memory expands over time as more content is translated.

Benefits of Translation Memory

Translation memory improves efficiency and consistency in multilingual workflows.

• Reduces repetitive translation work
• Improves terminology consistency
• Speeds up translation workflows
• Lowers translation costs
• Preserves valuable translation assets

Memory in AI Translation Workflows

Translation memory is often combined with machine translation systems to improve translation quality. While machine translation generates new translations, translation memory ensures consistent reuse of approved content.

LILT’s AI-powered translation platform combines translation memory with adaptive machine translation and human feedback to help organizations scale high-quality multilingual content.